AltaySec · ATLAS 2026.07
AML.T0044

Full AI Model Access

C · Explainer (cannot be honestly simulated in-browser) AML.TA0000 · AI Model Access

Description

Adversaries may gain full "white-box" access to an AI model. This means the adversary has complete knowledge of the model architecture, its parameters, and class ontology. They may exfiltrate the model to Craft Adversarial Data and Verify Attack in an offline environment where it is hard to detect their behavior.

Honesty-tier rationale

An access precondition, not an attack. There is no model behavior to enact — delivered as concept explainer.

Source

← Back to the matrix